Delft University of Technology
Optimal System Design for a Solar Powered EV Charging Station
Chandra Mouli, Gautham Ram; Bauer, Pavol DOI
10.1109/ITEC.2018.8450083 Publication date
2018
Document Version Final published version Published in
2018 IEEE Transportation and Electrification Conference and Expo, ITEC 2018
Citation (APA)
Mouli, G. R. C., & Bauer, P. (2018). Optimal System Design for a Solar Powered EV Charging Station. In 2018 IEEE Transportation and Electrification Conference and Expo, ITEC 2018 (pp. 1094-1099). [8450083] Danvers: IEEE. https://doi.org/10.1109/ITEC.2018.8450083
Important note
To cite this publication, please use the final published version (if applicable). Please check the document version above.
Copyright
Other than for strictly personal use, it is not permitted to download, forward or distribute the text or part of it, without the consent of the author(s) and/or copyright holder(s), unless the work is under an open content license such as Creative Commons. Takedown policy
Please contact us and provide details if you believe this document breaches copyrights. We will remove access to the work immediately and investigate your claim.
This work is downloaded from Delft University of Technology.
Green Open Access added to TU Delft Institutional Repository
‘You share, we take care!’ – Taverne project
Optimal System Design for a Solar Powered EV
Charging Station
Gautham Ram Chandra Mouli, Pavol Bauer
Dept. of Electrical Sustainable Energy, Delft University of Technology, Netherlands
G.R.Chandramouli@tudelft.nl, P.Bauer@tudelft.nl
Abstract- Charging electric vehicles (EV) from photovoltaic
(PV) panels provides a sustainable mode of transportation. In order to reduce the net costs of charging EV from PV and the grid, the PV generation and/or the EV charging can be controlled based on the energy prices in the grid. The traditional approach to designing the solar system for EV charging is to maximize the energy yield. In this paper, an alternate approach to PV system design is proposed by which the PV panels are orientated so as to maximize the PV revenue. This technique is compared with that of reducing the net costs by smart charging of the EV based on energy prices. Two case studies for Netherlands and Texas are done to compare the PV energy generated and the net cost of EV charging from PV based on the two techniques.
I. INTRODUCTION
Electric vehicles (EVs) are considered to be a clean mode of transportation as they have zero tail-pipe emissions. However, electric vehicles are only sustainable if the electricity used to charge them comes from sustainable sources. Unfortunately, the current electricity grid continues to be largely powered by fossil fuels, dominated by coal and natural gas [1]. So, when EVs are charged from such a grid, it results in indirect emissions at the power plants [2], [3].
A. Charging electric cars from photovoltaic panels
Charging of EVs from photovoltaic panels (PV) provides a distributed and sustainable method for powering electric vehicles [4]–[8]. There are several benefits to charging EV from PV such as,
Reduced demand on the grid as the EV charging power is locally generated from PV [5]
EV battery can be used as energy storage for the PV reduced cost of EV charging and reduced impact of
changes in feed-in-tariffs [3]
Fig. 1 shows an electric vehicle charging station that is powered by solar panels installed on the top of the building and as a solar carport at a workplace. Since EV battery and PV are both fundamentally DC by nature, an integrated charger can be used for direct DC charging of EV from PV as shown in Fig. 2 [7]–[9]. The power balance equation for the charger including the energy conversion losses will be
(1) where is the power drawn or fed to the grid, is the generated PV power, is the EV charging power. In the ideal case, =0 and the PV generation exactly matches
with the EV charging demand. However, this is hardly the case in practice due to the diurnal and seasonal variation in solar generation. The solution to matching the PV generation and EV charging is to either design the PV system or control the EV charging so that closely matches with .
B. Literature review
Firstly, the EV charging can be controlled to match the PV generation; a method commonly referred to as smart charging [10]–[14]. In case smart charging, linear programming, non-linear programming and fuzzy logic can be used for optimizing the EV charging profile to closing match the PV generation and the periods of low energy prices [13], [10], [11]. Solar forecasting can help in improving the optimization, for example, the online short-term solar power forecasting [15], the autoregressive integrated moving average (ARIMA) models or any of the methods listed in [16] can be used.
The second approach is to optimize the PV system for meeting the EV charging demand. A simple but expensive way to do this is to use a dual-axis solar tracker, to get the
Fig. 1. Impression of a solar powered EV charging station
Fig. 2. Schematic of 10kW grid-connected solar EV charger
1094 978-1-5386-3048-8/18/$31.00 ©2018 IEEE
maximum solar energy yield [17]. A cheaper approach is to install the PV system with a fixed orientation such that the orientation of the PV panels (tilt and azimuth of the modules) is so as to maximize the energy yield or match with the load or to increase the PV revenues [18]–[20]. While the first two methods are excellent from an energy point of view, it is not necessarily optimal from an economic perspective. This is because PV power is generally maximum in the afternoon, which is not always the time when the energy prices are high. So, if the net charging cost of EV from PV has to be reduced, it is important to orient the PV panels so as to increase the PV revenues.
C. Contribution
In case of charging EV from PV, the net cost is dependent on the cost of energy drawn from the grid given by
(2)
where is the energy price for the time period t. If is fixed, then maximising leads to lowering of the net cost. However, if varies with time, then it is important to maximize the PV revenue, or minimize the EV charging cost, , in order to reduce the net costs of charging EV from PV.
Hence, the aim of this paper is to implement and compare two techniques to reduce the net costs of EV charging from PV. First is to optimally design the PV system (tilt and azimuth as shown in Fig. 3) in order to maximize the PV revenues , instead of the traditional approach of maximizing energy yield, . The PV system is
therefore designed to generate maximum energy at times of high energy prices and vice versa, thereby reducing the net cost. The second technique is to implement smart charging by controlling the EV charging power as to reduce the EV charging cost, and the net cost, . Two cases namely Netherlands and Texas are considered for comparing the two techniques. The choice is because Netherlands and Texas are different in terms of solar irradiance, temperature and energy prices and hence the comparison is expected to highlight the influence of these parameters.
II. SYSTEM PARAMETERS
For the analysis, a grid-connected solar charging station with a 10kW PV array is considered, as shown in Fig. 2. The 10kWp PV array is composed of 30 modules (5 strings of 6
series connected modules) of Sun power E20-327 modules rated at 327W, whose specifications are shown in TABLE I . For the case of the Netherlands, meteorological data for solar irradiance and temperature from the Dutch Meteorological Institute (KNMI) for Cabauw for 2015 is used, which has a resolution of 1 minute. In case of Texas, meteorological data with 1 min resolution is extracted from the Meteonorm software for the city of Austin, Texas.
The focus of this work is on workplace charging of EV from PV. This is because workplaces are ideal for solar EV charging as the employees’ cars are parked for around 8 hours in the day when the sun is shining. With the long parking times, low charging powers are sufficient to provide adequate energy to the EV battery. For this study, it is assumed that employees are at the workplace from 9AM-5PM and the EVs are charged to with 29.6kWh of energy daily. This corresponds to an annual EV demand of 10804 kWh.
III. PV SYSTEM MODELLING
A. Estimating the PV module irradiance
In order to estimate the PV energy generation for different orientation, a model of the PV system is built in MATLAB. To estimate the solar irradiance on a module (Sm) with an
azimuth (Am) and tilt angle (θm), an estimation of the position
of the sun is required as shown in Fig. 3. A solar position calculator is hence built by which the azimuth (As) and
altitude (as) of the sun throughout the year at any location can
be determined [21]. The azimuth angle Am, As can range from
0° to 360° and the sign convention is 0° for North (N) and 180° for South(S). Similarly, θm, as can range from 0° to 90°.
With the sun’s position, the irradiance on a panel, with a specific orientation (Am, θm) can be estimated using the
geometric models in [18], [22] and the isotropic sky diffused model [22], [23]:
(3)
1 /2 (4)
(5)
where SDHI is the Diffuse Horizontal Irradiance (DHI), SDNI is
the Direct Normal Irradiance (DNI) and SDNI
m, SDHIm are the
components of DNI and DHI which is incident on the panel. From the above equation, we can see that the irradiance on
S N W E θm Zenith Am Sun as As
Fig. 3. Orientation of the PV panel is defined by azimuth angle Am
(measured from North) and module tilt angle θm (measured from horizontal
surface)
TABLE I
PARAMETERS OF SUN POWER E20-327 MODULE
Quantity Value
Area of module (Apv) 1.63 m2
Nominal Power (Pr) 327 W
Avg. Panel Efficiency (η) 20.4% Rated Voltage (Vmpp) 54.7 V
Rated Current (Impp) 5.98 A
Open-Circuit Voltage (Voc) 64.9 V
Short-Circuit Current (Isc) 6.46 A
Nominal Operating Cell Temperature (TNOCT) 45° C +/– 2 °C
Power Temp Coefficient (λ) –0.38% / oC
the panel can be influenced by changing the module azimuth (Am) and tilt angle (θm). Typically, the module tilt angle (θm)
can be used to control the seasonal variation in the solar generation as the sun has a high altitude in summer and much lower altitude in summer. This means that a high module tilt increases the winter solar generation while a lower module tilt increases the summer generation. Similarly, the module azimuth angle (Am) can be used to control the diurnal
variation in the solar generation by facing the modules east to increase the generation in the morning and modules west to increase the generation in the evening.
B. PV power and energy output
In order to estimate the power of a PV array based on the panel irradiance , it is important to consider the ambient temperature. The E20-327 PV module is rated for 327W at the STC ambient temperature of 25º. For other ambient temperatures (Ta), the PV array output power ( ) can be
estimated using [19], [24], where Tcell is the temperature of
the PV cells and is the number of modules in the array:
20
800 (6)
1 25
1000 (7)
IV. PVORIENTATION FOR MAXIMUM ENERGY Based on equations (3)-(7), the annual energy yield for different module tilt and azimuth is estimated for the case of Netherlands (NL) and Texas (TX) as shown in Fig. 4 and Fig.
5. The azimuth and tilt of the modules are varied in steps of 5° and 2°, respectively. The values for DHI, DNI, Ta are
obtained for the year 2015 from the KNMI for Cabauw, Netherlands (51.971°N, 4.927°E) and from the Meteonorm software for Austin, Texas (30.155°N, 97.445°W) that have a data resolution of 1 min.
A. Cabauw, Netherlands scenario
It can be seen for NL in Fig. 4 that the maximum annual yield of 11,593 kWh is obtained for south-facing panels with Am=185°, θm=28°. On the other, the lowest annual yield of
3,238.5 kWh is obtained for North facing panels Am=0°,
θm=90°. This shows that the annual yield can reduce by a
factor 3.58 depending on the orientation of the panels as summarized in TABLE II. The annual yield gradually reduces as the tilt is increased or decreased from 28° and/or the azimuth of the panel is set away from the southern direction.
To further elaborate the effect of orientation, Fig. 6 shows the power output over one day of the 10kW PV system for day 155 of the year 2015 for the south, west, and south-east orientation, with the same tilt angle of 28°. It can be seen Fig. 4. Annual energy yield of a 10kW PV system in the Netherlands for
different tilt and azimuth of modules
Fig. 5. Annual energy yield of a 10kW PV system in the Texas for different tilt and azimuth of modules
Fig. 6. Power generated by 10kW PV system for a summer day (Day 155 of year 2015) for south-east, south-west and south facing modules, all with a tilt angle of 28°
Fig. 7. Power generated by 10kW PV system for a spring day (Day 80 of year 2015) for south-east, south-west and south facing modules, all with a tilt angle of 28°. The legend is same as Fig. 6
TABLE II
PVORIENTATION FOR MAXIMUM & MINIMUM ANNUAL ENERGY YIELD
Annual yield PV Yield (kWh) Am (°) θm (°)
NL Maximum yield 11,593 185 28 Minimum yield 3,238 0 90 TX Maximum yield 15,654 175 18 Minimum yield 3,724 0 90
how the east and west facing panels facilities the increased generation of power in the morning and evening hours of the day, respectively. However, this only occurs on days with sufficiently high DNI. On a cloudy day with high DHI and little or no DNI, the effect of the module azimuth on the output is close to zero, for the same tilt angle of the panels. This is shown in Fig. 7 for day 80 of the year 2015 where panels with south, south-west and south-east orientation with the same tilt angle of 28° have nearly the same power output as well. Therefore, Fig. 6 and Fig. 7 together show both the potential and the limitation of controlling the output PV power by controlling the azimuth of the module. The module tilt, on the other hand, facilitates the control of the output power over the seasons of the year (not shown in the figure).
Austin, Texas scenario
In case of TX, Fig. 5 shows the annual yield of the PV system for different azimuth and tilt angles of the modules. The maximum yield of the PV system is 15,654 kWh which is 35% higher than the annual yield for the Netherlands case. The orientation for maximum yield is Am=175° and θm=18°,
and the lower tilt angle can be explained by the fact that Texas is at a lower latitude than the Netherlands. The minimum annual yield is 3,724 kWh, and it occurs when the orientation of the module is Am=0°, θm=90°, i.e., a north
facing module that is oriented perpendicular to the ground. V. PV ORIENTATION FOR MINIMUM NET COST In the previous section, the PV orientation for maximum energy yield was determined. In order to use the PV system for EV charging and to reduce the net cost of charging with variable energy prices, it is important to orient the modules so as to maximize the revenue.
A. Energy prices
For estimating the net cost , day-ahead market (DAM) energy prices for 2015 from the Amsterdam Energy Exchange (APX) and Electric Reliability Council of Texas (ERCOT) are used for the Netherlands and Texas case, respectively as shown in Fig. 8. A wide variation in the costs can be seen between the months and between the Texas and Netherlands case. The annual average price for APX and ERCOT was found to be 2.64c€/kWh and 3.99c$/kWh, respectively.
Fig. 9 shows the average electricity price over a 24h period for ERCOT and APX. It is interesting to note that the prices peak in the morning and evening and dip in the middle of the day but the nature of this variation is very different for NL and TX. For NL, the morning and evening peaks are very close in price, and the dip in prices occurs in the afternoon around 4PM. On the other hand for TX, the evening peak prices occurs around 5PM and are much higher than the morning peak and those of the rest of the day.
B. Optimal orientation for minimum cost: NL
Based on the method in the previous section to estimate annual yield for different orientation, the net cost for EV
charging from PV are estimated based on (2). The EV is assumed to be charged at a fixed charging power of 3.7kW from 9AM to 5PM with no smart charging.
Fig. 10 shows the net cost for different combinations of module azimuth and tilt for the Netherlands and Texas scenario and several interesting observations can be made. Firstly, for the Netherlands case, the orientation resulting in maximum PV revenue, ∑ and the lowest net cost, =25.21€ corresponds to Am=180°, θm=28°, as shown in
TABLE II. This orientation is not very different from the orientation for maximum yield (Am=185°, θm=28°). Secondly,
the shape of the contour plot in Fig. 10(a) closely matches the contour plot of Fig. 4. Thirdly, the net costs were found to be maximum for Am=0°, θm=90° with =391.65€, which is
much higher than the minimum net cost value of =25.21€. These observations point to the conclusion that the orientation for maximum yield results in maximum revenue as well for the Netherlands case considered. This is explained by the fact that APX 2015 energy prices has the first morning peak close to the afternoon in Fig. 9 when the PV generation is maximum. So orienting the panel to the south-east or south-west has the double disadvantage of lower energy yield as seen in Fig. 4 and lower revenue as seen in Fig. 9.
C. Optimal orientation for minimum cost: TX
On the other hand for the Texas case, the orientation resulting in the minimum net cost of =(-130.45$) corresponds to Am=225°, θm=20°. This orientation is facing
westward by 50° with a marginally higher tilt of 2° when Fig. 8. Hourly day-ahead market prices from APX and ERCOT for 2015. For scale, values above 120$/MWh are not shown.
Fig. 9. Hourly day-ahead market prices from APX and ERCOT for 2015 averaged over a 24 hour period
Pri
ce
compared to the orientation for maximum yield (Am=175°,
θm=18°) as shown in TABLE II. Secondly, the net costs for
the TX case are negative, mainly driven by the fact that 10kW PV system generates more energy compared to the NL case. Thirdly, even though the orientation for “Minimum net cost” has 9.73% lower annual yield than the orientation for “Maximum yield”, it still delivers 2.54% higher PV revenues and 10.97% lower net costs. Finally, the shape of the contour plot in Fig. 10(b) is very different from the contour plot of Fig. 5.
These observations show the influence of PV orientation on PV revenues when energy prices are considered. This is because the ERCOT 2015 energy prices, on an average, rise continuously from 9AM and peak at 5PM. This causes westward facing panels that generate more energy in the afternoon benefit from the higher energy prices.
VI. SMART CHARGING FOR MINIMUM NET COST
A. Charging algorithm
In this section, the aim is to implement smart charging of the EV so as to minimize the EV charging costs, ∑
and the EV-PV net costs over one day, . By doing so the optimized net costs over the entire year, can be minimized for the smart charging scenario. The PV have panels have the same orientation as the orientation for maximum yield as seen in section IV. Smart charging is done based on the energy prices using the formulation:
Minimize: ∑ (8) 0 < < 10kW ∀ t (9) ∑ = 30kWh ∀ t (10)
=0 ∀ t<9AM, t>5PM (11) ∑ d=1 to 365 (12) Linear programming in MATLAB is used to implement the optimization over a 24 hour period from t=00:00h to 23:59h for each day of 2015 using APX and ERCOT energy prices for the NL and TX case, respectively. Since the resolution of the PV data is 1 min, =1min as well. TABLE IV shows the annual EV charging costs and the annual net costs for the smart charging (SC) scenario for the NL and TX case. The net costs are compared in TABLE IV with the net costs for the scenario with economically optimal orientated PV (PVO) with fixed EV charging power of 3.7kW taken from TABLE II.
It can be clearly seen in TABLE IV that smart charging of EV based on energy prices results in much lower net costs than those obtained from optimally orientating the PV based
on prices. For the NL and TX case, optimal PV orientation increases the PV revenues only by a factor of 0.06% and 2.5%, respectively. On the other hand, smart charging of EV is much better and reduces the annual EV charging costs by 20.1% and 36.34% for NL and TX case, respectively.
B. Implementation aspects
Although two cases for Netherlands and Texas have been simulated here with day-ahead market prices, the method provided in this paper can be applied to different locations, and real-time market or intraday market prices can be used as well. The actual increase in PV revenues and reduction in net costs will vary on a case by case basis depending on the meteorological conditions, EV charging profile and the nature of the energy prices.
Secondly, besides the PV model used here, other PV models can be used as well, especially for the DHI and the thermal modeling of the PV. Thirdly, the losses in the power
(a) Net cost of EV charging from PV for NL (€)
(b) Net cost of EV charging from PV for TX ($)
Fig. 10. Net cost of EV charging from PV for different tilt and azimuth of 10kW PV system for (a) Netherlands and (b) Texas scenario
TABLE IV
COMPARISON OF NET COSTS FOR SMART CHARGING (SC) AND OPTIMALLY ORIENTED PV(PVO) FOR NL AND TX
PV Yield (kWh) PV revenue (€ or $) Charging EV Cost (€ or $) Net cost (€ or $) Am (°) θm (°) NL 11,593.4 506.58 424.95 -81.63 185 28 SC 11,592.9 506.89 532.10 25.21 180 28 PVO TX 15,654 506.76 247.73 -259.02 175 18 SC 15,245 519.65 389.20 -130.45 225 20 PVO TABLE III
PVORIENTATION FOR MAXIMUM & MINIMUM NET COST Annual PV Yield (kWh) Net cost (€ or $) PV revenue (€ or $) Am (°) θm (°) NL Max. yield 11,593.4 25.52 506.58 185 28 Min. net cost 11,592.9 25.21 506.89 180 28 Max. net cost 3,238.5 391.65 140.45 0 90 TX
Max. yield 15,654 -117.55 506.76 175 18 Min. net cost 15,245 -130.45 519.65 225 20 Max. net cost 3,761 270.94 118.26 5 90
electronic converter for the PV and EV will marginally increase the net costs, and this aspect has been neglected in this work. Finally, full/partial shading of the PV panels due to nearby objects and buildings will reduce the PV output depending on their location and size. These factors are, however, beyond the scope of this work and can be considered in the future.
VII. CONCLUSIONS
This paper has shown that in a scenario with variable energy prices, there is a potential to orient the PV system and/or implement smart charging in such a way so as minimize the net cost of charging electric vehicles from solar. Installing the PV system based on the variation of electricity prices is contrary to the conventional approach of maximizing energy and has untapped potential for future EV-PV applications.
It was found that a PV system oriented with azimuth
Am=185°, tilt θm=28° and Am=175°, θm=18° results maximum
annual energy yield for the case of Cabauw, Netherlands, and Austin, Texas, respectively. If electricity prices for 2015 from APX and ERCOT were considered, then the optimal PV orientation for the minimum net cost for EV charging from PV was found to be Am=180°, θm=28°, and Am=225°, θm=20°
for the Netherlands and Texas case, respectively. Thus, for the Netherlands case, the influence of 2015 APX prices was minimal, and the orientation for maximum yield and for minimum net costs was nearly the same. On the other hand for the Texas case, the ERCOT prices showed a trend to increase in the afternoon thus encouraging the solar panels to be oriented to the west so as to increase the PV revenue.
In the case of smart charging of EV based on energy prices, the annual EV charging costs were reduced by 20.1% and 36.34% for NL and TX case, respectively when compared to charging at a fixed power. Further, this reduction in EV charging costs was much higher than the marginal increase in PV revenues of 0.06% and 2.5% obtained by orienting the PV based on energy prices for NL and TX, respectively.
ACKNOWLEDGMENT
The authors would like to acknowledge J. H. Castro Barreto, O. Isabella, V. V. Ashok, C. Onwudinanti, G. Julian from the Delft University of Technology for their support in this research work.
REFERENCES
[1] “Efficiencies and CO2 emissions from electricity production in the
Netherlands, 2012 update,” Cent. Bur. Stat. - Netherlands, 2014. [2] O. Van Vliet, A. S. Brouwer, T. Kuramochi, M. Van Den Broek, and A.
Faaij, “Energy use, cost and CO2 emissions of electric cars,” J. Power
Sources, vol. 196, no. 4, pp. 2298–2310, 2011.
[3] G. R. Chandra Mouli, M. Leendertse, V. Prasanth, P. Bauer, S. Silvester, S. van de Geer, and M. Zeman, “Economic and CO2 Emission Benefits of a Solar Powered Electric Vehicle Charging Station for Workplaces in the Netherlands,” in IEEE Transportation Electrification Conference and Expo (ITEC), 2016, pp. 1–7.
[4] D. P. Birnie, “Solar-to-vehicle (S2V) systems for powering commuters of the future,” J. Power Sources, vol. 186, pp. 539–542, Jan. 2009.
[5] G. R. Chandra Mouli, P. Bauer, and M. Zeman, “System design for a solar powered electric vehicle charging station for workplaces,” Appl. Energy, vol. 168, pp. 434–443, Apr. 2016.
[6] G. R. C. Mouli, P. Venugopal, and P. Bauer, “Future of electric vehicle charging,” in International Symposium on Power Electronics (Ee), 2017, pp. 1–7.
[7] G. R. Chandra Mouli, J. H. Schijffelen, M. van den Heuvel, M. Kardolus, and P. Bauer, “A 10kW Solar-Powered Bidirectional EV Charger Compatible with Chademo and COMBO,” IEEE Trans. Power Electron., 2018.
[8] C. Hamilton, G. Gamboa, J. Elmes, R. Kerley, A. Arias, M. Pepper, J. Shen, and I. Batarseh, “System architecture of a modular direct-DC PV charging station for plug-in electric vehicles,” in Annual Conference on IEEE Industrial Electronics Society, 2010, pp. 2516–2520.
[9] G. R. Chandra Mouli, P. Bauer, and M. Zeman, “Comparison of system architecture and converter topology for a solar powered electric vehicle charging station,” in International Conference on Power Electronics and ECCE Asia (ICPE-ECCE Asia), 2015, pp. 1908–1915.
[10] G. R. Chandra Mouli, M. Kefayati, R. Baldick, and P. Bauer, “Integrated PV Charging of EV Fleet Based on Dynamic Prices, V2G and Offer of Reserves,” IEEE Trans. Smart Grids, Accept. Publ., 2017. [11] T. Ma and O. A. Mohammed, “Optimal Charging of Plug-in Electric
Vehicles for a Car-Park Infrastructure,” IEEE Trans. Ind. Appl., vol. 50, no. 4, pp. 2323–2330, Jul. 2014.
[12] B. V. E. Bakolas, P. Bauer, and D. Prins, “Testing of Smart Charging Controller for dynamic charging from solar panels,” in IEEE Transportation Electrification Conference and Expo, 2014, pp. 1–4. [13] G. R. C. Mouli, J. Kaptein, P. Bauer, and M. Zeman, “Implementation
of dynamic charging and V2G using Chademo and CCS/Combo DC charging standard,” in IEEE Transportation Electrification Conference and Expo (ITEC), 2016, pp. 1–6.
[14] D. van der Meer, G. R. Chandra Mouli, G. Morales-Espana, L. Ramirez Elizondo, and P. Bauer, “Energy Management System with PV Power Forecast to Optimally Charge EVs at the Workplace,” IEEE Trans. Ind. Informatics, vol. 14, no. 1, pp. 311–320, 2018.
[15] P. Bacher, H. Madsen, and H. A. Nielsen, “Online short-term solar power forecasting,” Sol. Energy, vol. 83, no. 10, pp. 1772–1783, Oct. 2009.
[16] M. Diagne, M. David, P. Lauret, J. Boland, and N. Schmutz, “Review of solar irradiance forecasting methods and a proposition for small-scale insular grids,” Renew. Sustain. Energy Rev., vol. 27, pp. 65–76, Nov. 2013.
[17] R. Eke, A. S.-S. Energy, and undefined 2012, “Performance comparison of a double-axis sun tracking versus fixed PV system,” Elsevier.
[18] V. V. Ashok, C. Onwudinanti, G. R. Chandra Mouli, and P. Bauer, “Matching PV Array Output With Residential and Office Load by Optimization of Array Orientation,” in PowerTech (POWERTECH), 2015 IEEE Eindhoven, 2015, pp. 1–6.
[19] O. Isabella, G. G. Nair, A. Tozzi, J. H. Castro Barreto, G. R. Chandra Mouli, F. Lantsheer, S. van Berkel, and M. Zeman, “Comprehensive modelling and sizing of PV systems from location to load,” MRS Proc., vol. 1771, pp. 1–7, Apr. 2015.
[20] H. M. S. Hussein, G. E. Ahmad, and H. H. El-Ghetany, “Performance evaluation of photovoltaic modules at different tilt angles and orientations,” Energy Convers. Manag., vol. 45, no. 15–16, pp. 2441– 2452, 2004.
[21] J. J. Michalsky, “The Astronomical Almanac’s algorithm for approximate solar position (1950–2050),” Sol. Energy, vol. 40, no. 3, pp. 227–235, Jan. 1988.
[22] P. P. G. Loutzenhiser, H. Manz, C. Felsmann, P. A. Strachan, T. Frank, and G. M. Maxwell, “Empirical validation of models to compute solar irradiance on inclined surfaces for building energy simulation,” Sol. Energy, vol. 81, no. 2, pp. 254–267, Feb. 2007.
[23] T. M. M. Klucher, “Evaluation of models to predict insolation on tilted surfaces,” Sol. Energy, vol. 23, no. 2, pp. 111–114, Jan. 1979. [24] E. Skoplaki and J. A. Palyvos, “On the temperature dependence of
photovoltaic module electrical performance: A review of efficiency/power correlations,” Sol. Energy, vol. 83, no. 5, pp. 614– 624, 2009.
